AverageRankScore: Average-based rank score

In mlDNA: Machine Learning-based Differential Network Analysis of Transcriptome Data

Description

This function calculates the activities of pathways in the whole genome with the average-based rank scoring algorithm (Yang, et al., 2011; Ma and Wang, 2013). This rank-based statistics is robust for directly comparing the activities of pathways with different gene numbers under different experimental conditions, since it produces a normalized value with the consideration of gene number in the analyzed pathways and whole genomes.

Usage

AverageRankScore( featureMat, selGenes )

Arguments

featureMat	a numeric matrix recording the expression levels or changes of all genes in the genome at given conditions.
selGenes	a character vector recording a set of genes in the analyzed pathway.

Value

value

a numeric vector recording the activities of interested genes (selGenes) at different conditions.

Author(s)

Chuang Ma, Xiangfeng Wang

References

[1] Huang Yang, Chao Cheng and Wei Zhang. Average rank-based score to measure deregulation of molecular pathway gene sets. PLoS One, 2011, 6(11): e27579.

[2] Chuang Ma, Xiangfeng Wang. Machine learning-based differential network analysis: a study of stress-responsive transcriptomes in Arabidopsis thaliana. 2013(Submitted).

Examples

  
## Not run: 

   ##generate expression feature matrix
   sampleVec1 <- c(1, 2, 3, 4, 5, 6)
   sampleVec2 <- c(1, 2, 3, 4, 5, 6)
   featureMat <- expFeatureMatrix( expMat1 = ControlExpMat, sampleVec1 = sampleVec1, 
                                   expMat2 = SaltExpMat, sampleVec2 = sampleVec2, 
                                   logTransformed = TRUE, base = 2,
                                   features =  "foldchange" )

   ##for an interested set of genes, the average-based rank score can be calculated:
   genes <- rownames(featureMat)[1:100]
   res <- AverageRankScore( featureMat = featureMat, selGenes = genes )


## End(Not run)

mlDNA documentation built on May 2, 2019, 2:15 p.m.